LLM
A neural network trained on huge amounts of text, capable of generating, summarising, and reasoning about language.
Why it matters
Large Language Models are the foundation of the current AI revolution. GPT-4, Claude, Gemini, Llama - these are all LLMs. Every AI writing tool, coding assistant, and chatbot you use is built on one.
Understanding LLMs helps you cut through marketing hype. When a tool claims to be "AI-powered," knowing which LLM it uses and how it is configured tells you what it can actually do.
How it works
4 stepsRelated terms
From the glossaryFrequently asked questions
Is every AI an LLM?+
No. LLMs are one type of AI model focused on text. Other AI models handle images, audio, video, or structured data prediction without being language models.
How big is a large language model?+
Size is measured in parameters, the numeric weights learned during training. Models range from a few million parameters for small local models to hundreds of billions for frontier models like GPT-4.
Can LLMs reason?+
They can simulate multi-step reasoning through chain-of-thought prompting, but current research debates whether this constitutes true reasoning or very sophisticated pattern matching.
See the tools that use it.
The fastest way to understand LLM is to see it inside real products. Browse hand-reviewed tools that put it to work, each one checked by a person before it was listed.
Browse hand-reviewed AI tools